Wireless Performance - more
As we saw in the NETGEAR R6300 retest, 2.4 GHz downlink performance tends to track very closely as signal levels drop. The only exception is the ASUS RT-AC66U as the plot clearly shows.
2.4 GHz Downlink Throughput vs. Attenuation
The 2.4 GHz uplink plot shows the reason for its relatively poor 2.4 GHz uplink showing. The 868L's throughput starts to fall off after only 21 dB of attenuation and stays under the WD My Net AC1300 the rest of the way down. It also failed to run tests with 60 and 63 dB of attenuation programmed.
2.4 GHz Uplink Throughput vs. Attenuation
The 5 GHz tests show draft 802.11ac performance with the downlink plot below showing the WD and 868L tracking fairly closely all the way down.
5 GHz Downlink Throughput vs. Attenuation
The 5 GHz uplink plot shows more separation between the WD and D-Link, with the latter besting both other routers only at the 0 dB test point, which is also used for the "Location A" equivalent.
5 GHz Uplink Throughput vs. Attenuation
D-Link is using its unusual cylindrical packaging to differentiate itself. But, aside from looking different, does it really help wireless performance? Our wireless test process includes four complete sets of tests, turning the router 90° each time. I do this so we don't penalize anyone for any near-field effects from our chamber test technique. I always use the best run of the four, but it has to be the same run for both up and downlink. If the average result is the best, I use that.
For the 2.4 GHz downlink tests, there's a pretty wild set of runs. The plot below shows the four test runs and the calculated average. The bolded 180° run (with the router connector panel facing the test chamber antennas) was the overall best for both up and downlink. The 270° run had better throughput at the lower levels, but started falling off way too early and then jumped up again.
Four test runs - 2.4 GHz downlink - D-Link DIR-868L
The uplink plot shown below was again best at the lower signal levels with the router in the 270° position. But since I use the same run for both up and downlink, I again used the 180° run.
Four test runs - 2.4 GHz uplink - D-Link DIR-868L
The upshot of this is that D-Link's unusual antenna arrangement is perhaps more directional than the cylindrical format might suggest—at least when the router and client are very close together. There's definitely something odd going on in the router's internal rate / power adjustment algorithms running downlink, however. The throughput decline at low attenuation levels occured in multiple runs at different orientations.
As the first AC1750 router using Broadcom's more powerful second-generation router SoC, the DIR-868L has a bit of an edge in storage sharing throughput over products using the first-gen BCM4706. The USB 3.0 port certainly contributes to that storage sharing throughput advantage.
But folks really pay the big bucks for AC1750 routers for the higher wireless throughput they are supposed to provide. And D-Link oddly has left customers on their own to find a solution from other vendors to get the most out of its flagship draft 802.11ac router. The DIR-868L's biggest disadvantage vs. its competition is its lack of a bridge mode and lack of a companion AC1300 class standalone bridge product.
Still, the DIR-868L as a whole is good enough to take the #2 spot for AC1750 class routers in our Router Ranker. The summary below shows that its main weakness is range, primarily in the 2.4 GHz band. With strong signals and in three out of four benchmarks, however, it's the current best of the AC1750 class bunch when it comes to wireless throughput as long as it has a good, strong signal.